Commercial training data from specialist providers. Listings show what each source advertises; licensing is confirmed with the provider before any agreement.
A benchmark text dataset developed for BFCL 4v function-calling evaluation based on Japanese.
A benchmark text dataset developed to evaluate long-context processing capabilities based on Japanese.
A multi-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Japanese.
A single-turn Chat Arena preference evaluation dataset compiled by collecting human preference judgments on AI model responses based on Japanese.
A Japanese general-domain multi-turn chat text dataset built by professional annotators for daily life and general conversation domains.
A Japanese handwritten document OCR image dataset with bounding-box annotations for multilingual document understanding, OCR model training, and layout-aware text…
Simulated IVR queries recorded over telephony
Simulated IVR queries recorded over telephony
A Japanese OCR image dataset built from collected images related to Japanese landscapes, supporting visual text recognition and domain-specific image understanding.
A Japanese OCR image dataset for real-world menu images in the food and restaurant domain, reviewed and annotated by professional annotators.
A Japanese multi-turn spoken dialogue dataset covering everyday, travel, and academic domains.
High quality simulated podcasts, recorded in-studio
High quality simulated podcasts, recorded in-studio
A Japanese real-world OCR image dataset built with OCR annotations for Japanese text detection, recognition, and Document AI model development.
Simulated IVR queries recorded over telephony
A Japanese OCR image dataset built from collected real-world storefront sign images, supporting sign text recognition and scene text understanding.
Simulated IVR queries recorded over telephony
A multilingual chain-of-thought reasoning dataset built from complex problems requiring step-by-step decomposition and coherent answer generation, with AI-generated…
A parallel translation corpus dataset built from Korean source texts and corresponding foreign-language translations across general domains, created by professional…
A text dataset for LLM Chain-of-Thought training developed on the basis of expert utterances in Japanese.
A benchmark text dataset developed for Arena evaluation based on Japanese.
A high-difficulty reasoning benchmark dataset constructed in accordance with global HLE standards based on Japanese encompassing complex reasoning problems at an…
A benchmark text dataset developed for Tau evaluation based on Japanese.
High quality simulated podcasts, recorded in-studio